Master data refers to facts that describe the core of entities, for example, an organization's employees, customers, suppliers, partners, organizations, products, materials, accounts, medical records, locations, and others. Such master data are of high value information that an organization uses repeatedly across many business processes. Enterprise Master Data Management by Allen Dreibelbis, Eberhard Hechler, Ivan Milman, Martin Oberhofer, Paul van Run and Dan Wolfson, IBM Press, 2008 provide background on master data management.
Managing master data faces challenges in that the data is usually scattered throughout the enterprise without consistent view of the master data. Fragmentation occurs as a result of the data being trapped inside enterprise resource planning (ERP), customer relationship management (CRM) and other commercially available off-the-shelf (COTS) packages. Factors such as mergers and acquisitions, experiments in new markets, decentralized businesses, and legal requirements across geographical boundaries also may contribute to fragmentation and inconsistency in master data.
Master data may be managed as objects and attributes, and by defining transactions over and access control to the objects and attributes. Data governance procedures may be also defined for functionalities such as conflict resolution, data import and data integration.
An MDM system or server should ensure consistent master information across transactional and analytical systems, address key issues such as data quality and consistency proactively rather than “after the fact” in the data warehouse, decouple master information from individual applications, become a central, application independent resource, and simplify ongoing integration tasks and new app development. An MDM system can be implemented with different styles. For instance, in “consolidation” style implementation, data is periodically replicated to the MDM server. In “register” style implementation, an MDM server knows where to find the data. In “transactional hub” style implementation, an MDM server becomes the system of record for master data. Applications should be appropriately architected to use this style of MDM implementation.
InfoSphere™ MDM is an MDM product from International Business Corporation (IBM), Armonk, N.Y. InfoSphere™ MDM product family includes Master Data Management Server with data model that include three domains (party, product, account), Master Information Hub that allows a user to make user's own domain and data models, and Master Data Manager for Product Information Management, which is a web-based collaborative authoring environment for a product domain in a data model. The party domain of the MDM Server manages the entirety of data related to parties such as customers, vendors, and suppliers, people and organization. The product domain of the MDM Server manages the definitions of products, category hierarchies, bundles, and equivalences. Its collection of products makes up a product catalog that is accessible throughout the enterprise. The account domain of the MDM Server manages business relationships and agreements with other parties, such as billing, claims and contracts. MDM functionalities include suspecting duplicate processing (also referred to as “data stewardship”) that prevents inadvertent creation of duplicate parties and products, for instance, using matching techniques; managing history of data modifications (also referred to as “point-in-time history”), which includes a full audit database that contains the full modification histories of all data objects and a set of query options for the audit database; keeping track of the source of all data and when it was added (also referred to as “source value”); maintain links to documents stored in a Content Management System (CMS) (also referred to as “document attachments”); and allowing administrators to define what elements and sub-elements users and user groups can see based on defined constraints (also referred to as “rules of visibility”).
To date, conventional use of master data management includes managing a single physical and logical MDM system for an entire enterprise, in which the scope of the applications and organizations of MDM is determined in the design stage. However, in many organizations, there may be requirements for multiple and multiple-level (hierarchical) MDM systems.